Financial Predictions based on Fusion Models-A Systematic Review

摘要

In this present economic era people want to generate money in terms of profit in a rapid time amount, which is proportional to their investment. This can be happened on making correct business decisions at right times. In general, the economy of any country is purely dependent on business transactions takes place all around the world. In all the sort of businesses the profit gained is proportional to investment, but one business platform which gives profit as multiples is the stock market. If a person selects the right stock - at right time, then the profits are earned. Hence, that the pair selection of the stock and a moment of corresponding stock plays a vital role in profit generation in stock market. The stock market is a very dynamic environment as it was very sensitive towards the each and every event occurred throughout the world. Even though it is very tough to estimate the next movement of any of the stock, many of the professionals are trying to predict the movements of stock market. They used several statistical methods, time series analysis models, machine learning forecasting techniques for the prediction task. The purpose of this study is to find out the suitable best model to predict the stock market, by studying various stock market prediction approaches, which combines different fields like data mining, machine learning and sentiment analysis. During the process of studying various techniques the common things noted are dataset to be formed with stock data and apply feasible preprocessing on it for making ready for further analysis. Apply machine learning algorithms and measure the prediction result as in terms performance measurements like accuracy, Mean Square Error etc. The final conclusions drawn from this survey are, forecasting is the very difficult task as the rapid changes in market movement takes place in stock market. And also for efficient and accurate predictions need to be considered various factors in real world. The efficient stock prediction act as great asset for several stock agencies and also used to overcome the problems of stock investors.